Patterns and Causes of Deviations in English Verbal Inflectional Suffixes among Thai ELF Learners

This study investigates the use of three verbal inflectional suffixes, i.e., the present tense -s, the past tense -ed, and the progressive -ing, among Thai ELF learners. It examines how they deviate from ENL norms and the causes of deviations are analysed. Data were taken from the academic writing of 116 English-major students at a university in Bangkok. The results showed that Thai ELF learners who have advanced and upper-intermediate level English knowledge and skills have acquired the ability to use these three suffixes, but they sometimes deviate from ENL norms. They tend to omit the -s ending when there is a long distance between the main subject and main verb, when there is a heavy subject containing a head and pre-/post-modifiers, and when the subject appears as a structurally complex category. They often omit the -ed ending when there are several past tense verbs in a sentence. They extend the use of the progressive aspect to talk about a general truth or habit which is typically expressed by the present simple tense in ENL. Results suggest that linguistic and functional causes are responsible for these deviations. Thai ELF learners use the zero forms of present and past tense verbs as a result of both syntactic complexity and the pragmatic motives of the efficiency of communication as well as the exploitation of redundancy. They use progressive verbs with general truths or habits due to the attractive form and meaning of this aspect and also the pragmatic motive of added prominence.


INTRODUCTION
The theoretical notion of "English as a Lingua Franca" (henceforth ELF) emerged in the second half of the 1990s.ELF is the form of English used as a contact language by people from varying lingua-cultural backgrounds (Firth 1996).Researchers using the ELF approach distinguish the form of ELF from English as a native language (ENL) and the pedagogic subject English as a foreign language (EFL).Traditionally, the study of English by nonnative speakers is largely based on ENL norms, with the goal of communicating effectively with native speakers; differences from ENL norms in all aspects are regarded as errors that result from learners' incomplete acquisition and when this takes place, teachers' correction and remediation are very much needed (Jenkins 2006).However, the aim of English study for an ELF speaker is to communicate with other non-native speakers, so technically, ENL norms should not be applied to set the linguistic agenda of ELF (Jenkins 2012).Instead, the norms of ELF are negotiated by its users for specific purposes by relying on their linguacultural resources and are derived from interactions involving efforts and adjustments from all parties (Jenkins 2009, Cogo 2010).
Based on this tenet of the theory, differences from ENL norms are not always signs of language incompetence, but can be indicative of emerging or potential features of ELF (Jenkins et al. 2011).That is, ELF speakers are viewed as language users in their own right who make use of English for their own purposes.According to Cogo and Dewey (2006) and Jenkins (2006), there are three criteria for determining whether features are emerging trends.First, they must be systematic in nature.Secondly, they must occur frequently and are produced by numerous speakers.Finally, they should not cause a breakdown in communication.Modifications that meet these criteria are regarded as "deviations" or "variants", which are a natural part of language contact and language change, and not as "errors", which are a result of the incomplete knowledge of English.
Despite the focus on linguistic form, there are fewer ELF studies on morphosyntax, compared to phonology and pragmatics.In fact, research at this level is also crucial to the understanding of ELF communication.Many morphosyntactic features are redundant and serve communal, not communicative, purposes, resulting in speakers avoiding them and deviating from ENL norms (Seidlhofer 2008).Furthermore, many studies have analysed ELF forms in Europe (such as France and Italy) and the outer circle of Asian countries (such as Singapore and India); however, studies on the expanding circle of Asian countries (such as Japan and Thailand) are far less common (Bolton 2008).In addition, ELF research has predominantly dealt with spoken interactions (Dewey 2014).Although there has been recent interest in written language, not many implications have been drawn (Jenkins et al. 2011).
Thailand is classified as an Asian country in the expanding circle firstly because it does not have a history of British colonisation and secondly, because English is used mainly for inter-cultural communication.There are just a few studies on Thai ELF learners and they mostly focus on socio-cultural aspects, not on the linguistic form of the language.To deal with this limitation and the other limitations of the previous literature as stated above, this study examines ELF's morphosyntactic system in the academic writing of Thai learners.Given that deviations in tenses and aspects are often found in L2 research (e.g., Ranta 2006, Stapa & Izahar 2010), the target features of the current study include the present tense -s, the past tense -ed, and the progressive aspect -ing.The objectives of the study are: (1) to examine how Thai ELF learners deviate from ENL norms in their use of the three verbal inflectional suffixes, and (2) to analyse the underlying causes of the emerging patterns.It is hypothesised that Thai ELF learners show some forms of deviations in the three suffixes owing to both linguistic and functional causes.The results help to provide more complete data for the establishment of the empirical description of ELF's linguistic form.

LITERATURE REVIEW
The review of the literature covers two areas: the background of the ELF approach and the role of ELF in Thailand.

THE BACKGROUND OF THE ELF APPROACH
Globalisation has brought about "the widening, deepening, and speeding up of worldwide interconnectedness in all aspects of contemporary social life" (Held et al. 1999, p. 2).The effect of globalisation is profound and far-reaching, causing significant changes in every aspect of life including language (Dewey 2007).As people inter-connect for various purposes on a global scale, there is a need for English to be a world-wide contact language.
When English has spread geographically and across domains, the backgrounds of its users have been integrated into the characteristic features of ELF interactions.As a result, ELF's common ground includes forms that it shares with ENL and also forms that differ from ENL and those that have arisen through contact between speakers across different geographical, linguistic, and cultural boundaries (Jenkins 2009).
Linguistic forms of ELF have been of interest to ELF scholars since the 1990s.This is because ELF has to be well-grounded in empirical description in order to gain acceptance as a legitimate version of English (Hülmbauer et al. 2008).ELF research has shown the patterns of change and linguistic fluidity emerging in the way English is transformed in lingua franca communication (Dewey 2007, Jenkins et al. 2011).Because it has been used by people with various linguistic backgrounds in multicultural contexts, ELF has undergone many deviations and innovations, and it is marked by a number of linguistic features that make it distinct from ENL and other English varieties on all levels, including phonology, pragmatics, and also morphosyntax.

THE ROLE OF ELF IN THAILAND
Almost 100% of the population in Thailand speaks the Thai language (National Identity Board 2000).Although Thailand has no official second language, English forms the de facto second language; it is a compulsory subject in schools and higher education, and is used in a wide range of domains such as business, tourism, the internet, advertising, and scientific and technological transfer (Foley 2005, Baker 2012).Thailand is also a member of ASEAN (the Association of Southeast Asian Nations), in which English is chosen as the official working language.Thus, English has an increasingly important role in the country.It is a lingua franca that links Thailand culturally, intellectually, and commercially with other ASEAN members and the rest of the world (Baker 2012).Given that English has such status, some forms of ELF used by Thai speakers should develop.A study on Thai learners not only reveals salient features particularly associated with this group of speakers, but also sheds light on the nature of ELF communication.

METHODOLOGY
This study employed a qualitative approach by examining the academic writing of Thai ELF learners and analysing how their use of the three verbal inflectional suffixes deviates from ENL norms and the causes of these deviations.

PARTICIPANTS
According to Cogo and Dewey (2006) and Jenkins (2009), although ELF interactions include users at all proficiency levels, its description and codification are drawn from communication between proficient users so that systematic emerging patterns can be identified.Accordingly, the target population of the study was advanced and upper-intermediate Thai ELF users who have received formal instruction in English and have been schooled to conform to ENL norms for several years.
A purposive sampling method was used to select the representatives of the population.The participants were Thai ELF learners who met the following language criteria.First, they had to be undergraduate and graduate students who majored in English at a university in Bangkok and had been in the programme for more than one year.They were targeted since they had studied the four skills of English -listening, speaking, reading, and writingextensively, especially during the period of their study at the university.Secondly, they had to get an average grade of over 3.20 for all English classes taken at the university to guarantee that they were at the advanced or upper-intermediate level of English knowledge and skills.100 undergraduate students took part in the study, including second-, third-, and fourth-year students from the English programme offered by the university.In addition, 16 graduate students also volunteered to participate in this study; they were either in the second or third year of their master's programme of English for International Communication at the same university.In total, 116 participants participated in this study.

INSTRUMENT
As Ranta (2006) suggested, the academic context is conducive to conduct ELF research.ELF speakers in academic settings use English as their working language regularly, hence, they tend to be experienced language users.In addition, the academic context is a demanding genre that requires learners to perform several kinds of challenging task, such as arguing a point, defending their view, and discussing a complex topic.Since managing this genre of language use is considered the highest point on an L2 attainment score, features found in the academic context can reflect new standards or emerging patterns.
Due to such advantages, data were taken from the academic writing of Thai ELF learners.The current study defines "academic writing" as a style of writing that is intended to inform (rather than to entertain), is written to address one particular topic, and is usually presented in three distinct sections in well-structured paragraphs (i.e., an introduction, a body, and a conclusion), with the following characteristics of language usage: formal and accurate, clear and precise, cohesive, and full of varied structures and word choices (Greetham 2001, Hamp-Lyons & Heasley 2006).
All participants attended two sessions of English writing tasks.They were instructed to write in a clear and logical way, using appropriate structure, vocabulary, and punctuation.They were also encouraged to use the standard style of academic writing which they had learned in their writing classes; that is, their essays should consist of three major components -an introduction (that includes a thesis statement), a body of 2-5 paragraphs (that discusses the main points and supporting details), and a concluding paragraph.Moreover, to ensure that their writing covers a wide range of issues, the first session was dedicated to writing on a casual or a less serious topic while the second one was dedicated to writing on an academic or a more serious topic.The two sessions took place on separate days so that the participants could concentrate thoroughly on their writing.

THAI ELF LEARNERS' USE OF THE VERBAL INFLECTIONAL SUFFIXES THE PRESENT TENSE SUFFIX
The number of sentences with marked and unmarked third person singular present tense verbs in the 232 essays of the 116 participants was 815.Since many sentences had more than one verb, present tense verbs outnumbered present tense sentences.In total, there were 974 third person singular present tense verbs; 846 were normally marked with the -s ending (86.86%) while only 128 showed zero marking (13.14%), as illustrated respectively in (1).
(1) a. First, the internet helps people communicate more easily.b.For example, WWE Raw, an American wrestling show, present the pictures of men beating each other with various techniques; or CSI: Miami, an American police procedural television series, show a lot of crime scenes and stories of criminals.
Based on the clear differences in the occurrence of third person singular present tense verbs in the inflectional form and those in the zero form, it can be stated that in general the participants have acquired the rule of the present tense suffix, which requires the addition of the suffix -s to a third person singular verb.An important question that followed was under what circumstances did the participants deviate from the morphosyntactic rule of the inflectional suffix -s?To deal with this question, three distinct characteristics of the participants' sentences with third person singular present tense verbs in two conditions, namely those with the -s ending and those with zero marking, were compared and analysed.
The three characteristics were (1) the distance between the main subjects and the main verbs, (2) the weight of the subjects, and (3) the syntactic categories of the subjects.

THE DISTANCE BETWEEN THE MAIN SUBJECTS AND THE MAIN VERBS
To check whether the distance between these two units had an effect on the participants' choice of third person singular present tense marking, the words between the main subject (the head of a subject) and the main verb (the head of a verb phrase) in each clause were counted.Lexical words (e.g., nouns, verbs) and non-lexical words (e.g., determiners, conjunctions) were all counted.Fixed or idiomatic expressions (e.g., compound words, phrasal verbs) and proper nouns (e.g., names of countries) behaved as one unit, so they were regarded as one single word.For example, in the sentence "To educate 1 them 2 properly seems to be the best reformation", there are two words between the main subject (educate) and the main verb (seems).
The results showed that the distance between the main subjects and main verbs affected the participants' choice of third person singular present tense marking.For clauses containing verbs with the regular -s suffix, the main subjects were mostly adjacent to the main verbs.The mean distance in these clauses was 1.18 words.In 491 clauses, the main subjects were placed immediately next to the main verbs (58.04%).In the other clauses where there were some words between these two linguistic units (41.96%), many of them -221 clauses -had only one or two intervening words (26.12%).On the other hand, for clauses containing verbs with zero marking, the main subjects were generally farther from the main verbs.The mean distance in these clauses was 2.70 words.The main subjects were put immediately next to the main verbs in only 40 clauses (31.25%).For other clauses, there were one or more words between these linguistic units (68.75%).To determine whether the weight of the subjects affected the participants' choice of third person singular present tense marking, the words that occurred in the subject position of each clause were counted.The procedure of word counting in the subject position was basically the same as the case of the subject-verb distance.Both lexical and non-lexical words were counted; idiomatic expressions and proper nouns were regarded as one word.For example, the subject " 1 the 2 scene 3 that 4 Elizabeth Gilbert 5 is 6 eating 7 spaghetti" consists of seven words.
The results showed that the weight of the subjects affected the participants' choice of third person singular present tense marking.For clauses containing verbs with the regular -s suffix, the subjects mostly appeared in a short simple form.The mean weight of all subjects was 2.06 words.648 subjects were made up of only one or two words (76.60%).On the other hand, for clauses containing verbs with zero marking, the subjects were generally heavier.The mean weight of all subjects was 2.81 words.There were 82 subjects consisting of one or two words (64.06%).Moreover, in all other heavier weight groups which contained more than two words in the subject position, the percentages of clauses with present tense verbs with zero marking were higher than those of clauses with present tense verbs marked with the suffix -s.To determine whether the syntactic categories of the subjects affected the participants' choice of third person singular present tense marking, the category of each subject was analysed.
The study differentiated words and phrases based on the traditional notion: a word consists of only one head and nothing else, whereas a phrase is made up of a head and other additional elements.For example, the one-word subject "water" was identified as a noun while the twoword subject "the water" was identified as a noun phrase.
The results indicated that the syntactic categories of the subjects affected the participants' choice of third person singular present tense marking.For clauses containing verbs with the regular -s suffix, the subjects occurred in ten syntactic categories.They appeared mostly as noun phrases (41.61%) and quite often as personal pronouns (30.50%).Interestingly, six of the ten categories had a very simple structure consisting of only one head word, either a noun or a pronoun; these categories included personal pronouns, nouns, proper nouns, indefinite pronouns, demonstrative pronouns, and interrogative pronouns.In total, approximately half of the subjects -424 subjects -belonged to these structurally simple categories (50.12%).Likewise, for clauses containing verbs with zero marking, the subjects occurred in the same ten syntactic categories, and they mostly appeared as noun phrases (57.03%).However, unlike clauses with verbs in the regular -s form, there were only 36 subjects of clauses with verbs in the zero form which belonged to the six structurally simple categories (28.13%).The number of sentences with past tense verbs marked and unmarked with the inflectional suffix in the 232 essays of the 116 participants was 1,081.Since many sentences had more than one verb, past tense verbs outnumbered past tense sentences.In total, there were 1,813 past tense verbs; 1,341 were normally marked with the -ed ending (73.97%) whereas only 472 showed past tense zero marking (26.03%), as illustrated respectively in (2).
(2) a.I once experienced this beautiful thing.b.However, I find it really amazing when I watch it the second time.The clear differences in the occurrence of past tense verbs in the inflectional form and those in the zero form suggest that in general the participants have acquired the rule of the past tense suffix, which requires the addition of the suffix -ed to a verb expressing a past meaning.What then were the circumstances in which the participants deviated from the morphosyntactic rule of the inflectional suffix -ed?A preliminary examination of the data showed that their past tense inflection seemed to rely on the number of past tense verbs in a sentence.Thus, the participants' sentences were compared and analysed in terms of the number of past tense verbs: (1) sentences with one past tense verb, (2) sentences with two or more past tense verbs, and (3) comparison of sentences with different numbers of past tense verbs.

SENTENCES WITH ONE PAST TENSE VERB
To check whether a sentence with one past tense verb had an effect on the participants' choice of past tense marking, the researcher separated sentences that had one past tense verb from other sentences.These sentences were divided into two groups.The first group included simple sentences with only one verb, which was a past tense verb.The second group consisted of sentences in various grammatical structures (such as simple, compound, and complex structures), which contained one past tense verb together with one or more verbs in different tenses.
The results showed that having one past tense verb in a sentence affected the participants' choice of past tense marking.Both groups showed high percentages of inflection marking.In the first group, there were 336 simple sentences with only one past tense verb; 248 sentences had a past tense verb marked with the -ed suffix (73.81%) while only 88 sentences contained a past tense verb with zero marking (26.19%).In the second group, there were 274 sentences with one past tense verb and one or more verbs in different tenses.Likewise, many of them -228 sentences -had a past tense verb marked with the -ed suffix (83.21%) whereas only 46 sentences contained a past tense verb with zero marking (16.79%).To check whether a sentence with two or more past tense verbs had an effect on the participants' choice of past tense marking, the researcher separated sentences that contained more than one past tense verb from other sentences.These sentences were divided into two groups.The first group included sentences that had two or more past tense verbs, all of which were marked with the -ed suffix.The second group consisted of sentences that had two or more past tense verbs, at least one of which showed past tense zero marking.
The results showed that having more than one past tense verb in a sentence affected the participants' choice of past tense marking.The total number of sentences that contained two or more past tense verbs was 471.There were only 266 sentences whose past tense verbs were all marked with the suffix -ed (56.48%).On the other hand, 205 sentences had at least one past tense verb that showed zero marking (43.52%).In this latter group, 79 sentences had all past tense verbs in the zero form (16.77%) while 126 sentences had past tense verbs in both the inflectional form and the zero form (26.75%).To check whether the increasing number of past tense verbs corresponded with the participants' higher tendency to leave out the past tense inflection, all 471 sentences that had more than one past tense verb were separated into three groups based on the specific number of their past tense verbs: two, three or four, and five or more past tense verbs.
The results showed that the different numbers of past tense verbs in a sentence affected the participants' choice of past tense marking.In the group of sentences with two past tense verbs, about half of them -175 sentences -had all past tense verbs marked with the -ed suffix (57.57%) while 129 sentences contained at least one past tense verb with zero marking (42.43%).The group of sentences with three or four past tense verbs showed similar proportions.While 83 sentences had all past tense verbs marked with the -ed suffix (57.64%), 61 sentences had at least one past tense verb with zero marking (42.36%).On the other hand, in the group of five or more past tense verbs, the proportion of all past tense verbs marked with the -ed suffix was noticeably lower.Only eight sentences contained all past tense verbs in the inflectional form (34.78%) whereas 15 sentences had one or more past tense verbs in the zero form (65.22%).The number of sentences with progressive verbs in the 232 essays of the 116 participants was 138.As some sentences had more than one verb, present progressive verbs outnumbered present progressive sentences.In total, there were 147 verbs that took the present progressive aspect.
Previous studies found that some groups of ELF speakers have developed a new function of the progressive aspect by using it for the purpose of gaining expressivity and explicitness in communication (Ranta 2006).To examine whether this deviation also occurred among the Thai participants, three characteristics of their sentences containing present progressive verbs were compared and analysed: (1) the meanings of the progressive aspect, (2) the weight of the verbs, and (3) the meanings of the verbs.

THE MEANINGS OF THE PROGRESSIVE ASPECT
Native speakers typically use the present progressive form to describe an ongoing situation.This basic meaning implies that a verb that can occur in this aspect possesses a meaning that is dynamic (rather than static) and the situation it describes happens with duration (rather than being instantaneous or punctual) (Huddleston & Pullum 2005).Apart from this meaning, native speakers use the progressive aspect to express five additional meanings; they use the progressive form with state verbs to make the verbs dynamic or emphasise the temporariness of situations, they use it with punctual verbs to extend the duration of situations, and they rely on this aspect to give a particular interpretation of their utterance, to reveal their affective stance, and to indicate a future plan (Downing & Locke 2006, Aarts 2011, Depraetere & Langford 2012).
To check the meanings of the present progressive in the participants' sentences, each sentence with a progressive verb was analysed in terms of the general meaning it was intended to express.The study identified the meaning of the progressive aspect based on the overall meaning of the target sentence with the progressive verb and also the meaning of contextual sentences.Any meaning beyond the meanings associated with the progressive aspect in ENL as described above was considered an extended, non-native use of this aspect.Note that the analysis in this part involved the meaning of the progressive aspect (the overall meaning of a sentence with a progressive verb), not the specific meaning of the verb itself.
The results indicated that the participants most frequently used the present progressive aspect to describe an ongoing situation.Almost half of the verbs, i.e., 72 verbs, took the present progressive aspect to convey that the situation in question was taking place at a particular moment (48.98%).Another meaning which was found quite often was a general truth or habit and 49 present progressive verbs were used to express this meaning (33.33%).Other meanings were much less common.Examples of the participants' use of the progressive aspect to convey an ongoing situation and a general truth or habit are provided respectively in (3).
(3) a.Currently, I am both studying and working as a volunteer teacher for children.b.Kids who are playing video games are stimulated to become more aggressive.Therefore, the participants most often used a verb in the progressive form to describe the typical meaning of an ongoing situation, and they extended the use of this aspect to talk about a general truth or habit.Given that these two meanings were contradictory (progressive vs. non-progressive), we assumed that the participants developed a new communicative function for the progressive aspect, which allowed them to use it with an ongoing situation and also a general truth or habit.To understand the communicative function of the progressive aspect, the form and meaning of verbs were then analysed.For each verb, only its infinitive form -excluding be and the -ing suffix, which were part of the present progressive structure -was examined to determine how many syllables it contained.For example, the present progressive verb phrase is singing has the infinitive form sing, so it was identified as a verb with one syllable.The results showed that the participants most frequently used the progressive aspect with short verbs.The mean weight of the 147 verbs was 1.40 syllables.Most of them -99 verbs -had only one syllable (67.35%), and there were 40 verbs with two syllables (27.21%).In addition, the study separated the 49 verbs that occurred in the progressive form to express the extended, non-native-like meaning of a general truth or habit.An analysis of their weight generally showed similar results.Most of them had a short form.The mean weight was 1.35 syllables; there were 34 verbs with one syllable (69.39%) and 13 verbs with two syllables (26.53%).All verbs were categorised according to their meaning.This study divided verbs largely based on the classification proposed by O' Grady (2001) and Downing and Locke (2006), which included various semantic classes of verbs such as verbs of volitional acts and verbs of communication.
The results showed that the participants' progressive verbs expressed various kinds of meanings and belonged to twelve semantic classes.Since the class of volitional acts involved a broad meaning including different kinds of volitional actions performed by the subject, 68 verbs conveyed this meaning (46.26%).Verbs in the other classes were found much less frequently.In addition, the study separated the 49 verbs that occurred in the progressive form to express the extended, non-native-like meaning of a general truth or habit.An analysis of their meanings generally showed similar results.They denoted various meanings and belonged to nine semantic classes; 28 verbs expressed volitional acts (57.14%) while verbs in the other classes were found far less frequently.Thus, the results involving the form and meaning of verbs suggested that the participants also used the present progressive aspect to gain expressivity and explicitness in communication.In terms of form, the progressive aspect, in which be and the suffix -ing were added to a main verb, made the verb longer and noticeable; as a result, the participants preferred using this aspect with a short verb to make it become outstanding and expressive.For meaning, the progressive aspect, in which some dynamic aspect and duration were added to a situation, made the meaning of a verb stand out; as a result, the participants used this aspect with different kinds of verbs from various semantic classes to render their meanings in more expressive and explicit ways.

CAUSES OF THAI ELF LEARNERS' DEVIATIONS
Based on the overall results about the participants' use of these three suffixes, we can summarise and make more general statements about Thai ELF learners as follows.Thai ELF learners at the advanced and upper-intermediate levels have generally acquired the ability to use the three inflectional suffixes correctly.They usually mark present and past tense verbs with the -s and -ed endings.They most often form the present progressive structure "be + Ving" to describe an ongoing situation.However, there are some situations in which they tend to deviate from the ENL norms.
For the two tense suffixes, their deviations mainly occur in the form of verbs, i.e., the -s omission of a third person singular present tense verb and the -ed omission of a past tense verb.An analysis of the results shows that linguistic causes are responsible for these deviations.In particular, Thai ELF learners may leave out the -s or -ed suffix when a sentence is syntactically complicated.In the case of the present tense, syntactic complexity happens in three major circumstances, involving the distance between the main subject and the main verb, the weight of the subject, and the syntactic category of the subject.When the main subject is separated from the main verb, when the subject is heavy, consisting of a head and pre-/post-modifiers, and when the subject belongs to a structurally complex category, Thai ELF learners tend to omit the suffix and use a verb in the zero form.As for the past tense, syntactic complexity is found in one major context involving the number of past tense verbs.When a sentence is structurally complicated containing two or more past tense verbs, Thai ELF learners often omit the suffix and use at least one of the verbs in the zero form.
Moreover, Thai ELF learners' deviations in the present and past tense suffixes are also due to functional causes.In written language, Thai ELF learners usually have time to think and organise their ideas into well-formed sentences.They add the -s or -ed suffix to a present or past tense verb to indicate the time frame of a situation and to make their sentence conform to the ENL norms.However, when the sentence becomes structurally complicated, they tend to deviate from the grammatical norms and put their focus on achieving effective communication.By paying more attention to various elements in a complicated structure and the overall meaning the elements are intended to convey, they sometimes leave out the present or past tense suffix and use a verb in the uninflected form.In other words, the deviations in the use of the present and past tense suffixes among Thai ELF learners are a result of both linguistic and functional causes.When a sentence is structurally complicated, Thai ELF learners turn their focus to making clear and effective communication, rather than grammatical norms, i.e., the addition of tense suffixes that are not vital to the overall meaning of the sentence.
The analysis of the functional causes of Thai ELF learners' deviations in the present and past tense suffixes demonstrates the interrelationship between pragmatics and morphosyntax.This supports Cogo and Dewey's (2006) study, which stated that pragmatic or functional motives often lead to changes and innovations in morphological and syntactic structures.According to their study, there are several kinds of pragmatic motives, such as efficiency of communication, increased explicitness, added prominence, exploitation of redundancy, and the reinforcement of proposition.In the case of the present and past tense suffixes, the pragmatic motives are effective communication and exploited redundancy.Thai ELF learners deviate from ENL norms by using the zero forms of present and past tense verbs as a result of their motivation to make effective communication and exploit the redundant features which do not affect the overall meaning of a sentence (i.e., the -s and -ed suffixes).
For the progressive aspect, their deviation involves the meaning of the progressive.Thai ELF learners extend the use of a verb taking the progressive aspect to talk about a general truth or habit.An analysis of the results suggests that this deviation is attributed to linguistic factors, i.e., the attention-catching form and the outstanding meaning of a progressive verb.According to Huddleston and Pullum (2005) and Ranta (2006), the source of the "attractiveness" of the progressive aspect resides in its form and meaning.Since the progressive aspect consists of the auxiliary be and a main verb with the -ing ending and it adds the dynamic aspect and duration to a situation, a verb that occurs in this aspect becomes longer and more noticeable and its meaning stands out.In other words, adding be and the -ing suffix gives the verb more prominence and salience in the speaker's utterance, and this helps to draw the interlocutor's attention (Ranta 2006).Thus, the results supported Ranta's (2006) corpus study of ELF speakers with different mother tongues (such as German, Russian, Swahili, and Chinese): Thai ELF learners have also realised the "attention-catchingness" and the communicative value of the progressive aspect, so they use it frequently to gain expressivity and explicitness in communication.They use a verb in the progressive form with different kinds of events to which they want to give prominence, including ongoing situations and also general truths or habits.
From this viewpoint, the deviation in the use of the progressive aspect is simultaneously due to linguistic and functional causes.Thai ELF learners extend the use of the progressive aspect to the non-native-like context of general truths or habits in order to make their utterance clearly understood.In other words, Thai ELF learners are aware of the attractive form and meaning of the progressive aspect and they extend its use for the purpose of gaining expressivity and explicitness in communication.
Like the tense suffixes, the functional cause of their deviation in the progressive form reflects the close relationship between pragmatics and morphosyntax.In this case, the pragmatic motive is added prominence.Thai ELF learners deviate from ENL norms by using the progressive aspect in the non-traditional context of general truths or habits as a result of their motivation to add prominence in communication.The results of this study indicated that although Thai ELF learners normally mark present and past tense verbs with the -s and -ed endings, they sometimes omit these suffixes and use verbs in the zero form.Given that for most cases they add the suffix to their present or past tense verb appropriately, this indicates that they have acquired the ability to use the two inflectional suffixes and that their deviations are not a result of ignorance or inadequate knowledge of ENL rules.Rather, they reflect some important characteristics and the nature of ELF communication.First, their deviations provide evidence that these two suffixes should not be included in ELF's core morphosyntactic features.Instead, they are redundant features which can be left out without causing a communication problem.The -s suffix carries information about the subject (a third person singular subject) and the tense (the present tense), yet all this information can be obtained by other means.Since English is not a pro-drop language, the information about the subject is revealed by the obligatory subject itself.Likewise, the information about the tense of a sentence can be relayed by lexical means (e.g., an adverb of time and other kinds of temporal modifier).The -ed suffix provides information about the tense (the past tense), but this is redundant because ELF users can indicate that a situation took place in the past with the help of words and phrases suggesting directly or indirectly the time of the situation.Because of the redundant information they carry, Thai ELF learners tend to omit the two suffixes in circumstances of syntactic complexity and turn their focus to making efficient communication.This viewpoint supports Seidlhofer (2008), who argued that many English grammatical features are so redundant that they are attentively protected for a communal, instead of communicative, purpose.
Second, their deviations in the use of these tense suffixes suggest that the most important objective of ELF interactions is to have clear and effective communication.In their writing, Thai ELF learners usually mark verbs in the present and past tenses with the -s anded suffixes to indicate the time frame of a situation and to make their sentence conform to the ENL norms.However, when they express a complex idea using a more complicated structure, they tend to be distracted from the standard form of English and turn their focus to effective communication.These results of the Thai ELF learners are consistent with several previous studies on ELF speakers from various lingua-cultural backgrounds; while traditional L2 pedagogy considers grammatical accuracy as an important factor in determining the success of English learning (Dewey 2014), conforming to L1 norms does not always ensure the efficiency of communication, so ELF users are found to adjust the norms and prioritise successful communication over grammatical rules (Jenkins et al. 2011).
As for the progressive, Thai ELF learners most often use the form "be + V-ing" to talk about an ongoing situation.They sometimes use the progressive with a state verb and a punctual verb to make the verb more dynamic and to extend the duration of the situation, respectively.These results show that they do not have a problem in acquiring the progressive; they recognise the typical and additional meanings of this aspect.This is not surprising because the progressive has been found to be the easiest verbal inflection to be acquired by L2 learners (Ranta 2006).Moreover, the results also suggest that their deviation in the progressive should not be viewed as an error resulting from the inadequate knowledge of the language, but it is an emerging pattern which reveals some characteristics and the nature of ELF interactions.
First, their deviation reflects the fluid and dynamic nature of ELF.ELF is a form of communication where users adjust the norms by drawing on a wide range of linguistic and cultural resources that are available in their repertoires to achieve mutual intelligibility.Thai ELF learners are aware of the attention-catching form and the outstanding meaning of the progressive aspect, so they make use of these linguistic resources to make their utterance clear and explicit.When they want to give more prominence to their verbs from various semantic classes, they put these verbs, particularly short verbs with one syllable, into the attractive progressive form.As a result, they have developed a new meaning of the progressive aspect.They use this aspect to talk about different kinds of situations to which they want to give prominence, including an ongoing situation and also a general truth or habit.
Second, like the tense suffixes, Thai ELF learners' deviation in the progressive reflects that the most important objective of ELF is to have successful communication.In ENL, the use of the progressive form has been extended due to stylistic reasons; speakers choose the progressive instead of the plain simple form to express their emotions, so it has developed various shades of meaning, including the interpretive progressive and the progressive of affect (Mair & Hundt 1995, cited in Ranta 2006).For Thai ELF learners, however, these two new meanings are not found in the data.Instead, their extended use of the present progressive, like some groups of ELF speakers, is due to expressive reasons (Ranta 2006).Thai ELF learners have developed a new function of the progressive; they rely on its attractiveness for the purpose of adding prominence (expressivity and explicitness), which finally leads to clear and successful communication.
In summary, the study of the three inflectional suffixes in the academic writing of Thai learners has shown that the -s/-ed omissions and the use of the progressive -ing with general truths or habits are emerging features in ELF, and both linguistic and functional causes underlie ELF deviations from the ENL norms in the morphosyntactic system.Moreover, the overall results of the study have provided some teaching implications.As Seidlhofer (2005) suggested, teachers of English for intercultural communication should make a distinction between features that tend to constitute the core of language and are crucial for intelligibility and (non-native) features that do not cause problems in communication.By acting on these insights, they do not need to spend much time on teaching "unproblematic variants" and can pay attention to forms and communicative strategies necessary for effective communication.As the present and past tense suffixes carry redundant information, they should not be treated as a big topic of teaching that requires extra practicing time for students who learn English mainly in international settings.On the other hand, the attractive form and meaning of the present progressive aspect should be pointed out because they demonstrate to students how to draw on their linguistic resources to achieve clear and effective communication.

TABLE 1 .
The distance between the main subjects and the main verbs

TABLE 2 .
The weight of the subjects

TABLE 3 .
The syntactic categories of the subjects

TABLE 4 .
Sentences with one past tense verb

TABLE 5 .
Sentences with two or more past tense verbs COMPARISON OF SENTENCES WITH DIFFERENT NUMBERS OF PAST TENSE VERBS

TABLE 6 .
Sentences with different numbers of past tense verbs

TABLE 7 .
The meanings of the present progressive aspect

TABLE 8 .
The weight of the verbs in the progressive aspect

TABLE 9 .
The meanings of the verbs in the progressive aspect

TABLE 10 .
Summary of deviations